Hemodynamic monitoring with Hypotension Prediction Index versus arterial waveform analysis alone and incidence of perioperative hypotension

Intraoperative hypotension is associated with increased morbidity and mortality. The Hypotension Prediction Index (HPI) is an advancement of the arterial waveform analysis to predict intraoperative hypotension minutes before episodes occur enabling preventive treatments. We tested the hypothesis that the HPI combined with a personalized treatment protocol reduces intraoperative hypotension when compared to arterial waveform analysis alone.


| INTRODUC TI ON
Intraoperative hypotension is a common adverse event during noncardiac surgery. 1,2 It is associated with an increased incidence of acute kidney injury, myocardial injury, neurological deficiencies, as well as an increased 30-day operative mortality. [3][4][5][6][7] Organ injuries were shown to be associated with the depth, frequency, and duration of hypotensive episodes. 4,8 Data indicate that a mean arterial pressure of 65 mmHg serves as a threshold to predict myocardial and kidney injury. [9][10][11] Therefore, continuous improvement in and development of new technologies to prevent hypotension remain important goals in modern anesthesiology.
More recently developed minimally invasive methods to decrease the incidence of hypotension comprise among others of the analysis of the arterial waveform. This advanced hemodynamic monitoring measures the arterial blood pressure continuously and calculates additionally cardiac output, stroke volume, stroke volume variation, and systemic vascular resistance. This deeper insight into hemodynamics enables goal-directed therapy (GDT) approaches; however, the given information only allows to react to events instead of preventing them. 12,13 Interestingly, recent data indicate a significant reduction in postoperative organ dysfunction by preventing intraoperative hypotension, suggesting a potential benefit of early intervention. 14 Therefore, the prediction of intraoperative hypotension and consequently its prevention by proactive treatment may show beneficial effects for patients. To achieve this goal, the Hypotension Prediction Index (HPI, Edwards Lifesciences Corp., Irvine, USA) as an advancement of the arterial waveform analysis (FloTrac, Edwards Lifesciences Corp., Irvine, USA) was developed based on a machine learning algorithm. The algorithm analyses physiological changes in the arterial waveform of the radial artery as prodromal signs of the imminent hypotension. Thus, the HPI aims to predict a hypotension up to 15 min prior to the event with a sensitivity of up to 88% and a specificity of up to 87%, enabling the anesthetist to intervene ahead and stabilize the blood pressure without an actual hypotensive event. 15,16 Both technologies can be used perioperatively as well as in intensive care medicine but are not approved for pediatric patients.
While the HPI algorithm is based on the FloTrac algorithm for the detection of individual pulse waves, we used a pragmatic study design to test the hypothesis that the application of the HPI in comparison to the FloTrac reduces incidence, duration, and severity of hypotensive events evaluated by time-weighted average of hypotension in moderate-or high-risk noncardiac surgical patients. Using a pragmatic study design, we included consecutively the first 100 patients who have been treated at Marien Hospital Herne with either the FloTrac sensor (n = 50) and the HPI sensor (AcumenIQ; n = 50) undergoing moderate-or high-risk abdominal surgery in urology, general surgery, and gynecology. The decision to use the advanced hemodynamic monitoring was based on standard operating procedures of the risk assessment for major abdominal surgery during premedication rounds by senior physicians. Inclusion criteria consisted of elective major abdominal surgery (e.g., cystectomy, pancreaticoduodenectomy, and cytoreductive surgery/HIPEC), age >18 years, anticipated duration of surgery >120 min, and need of invasive blood pressure monitoring using an arterial line. Exclusion criteria were patients not in sinus rhythm, ejection fraction <30%, severe aortic valve stenosis, emergency surgery, acute myocardial ischemia, anticipated duration of surgery <120 min, and contraindication for an arterial line. Patients consecutively enrolled in this study underwent surgery between 5 September 2019 and 5 August 2020.

| MATERIAL S AND ME THODS
No further criteria were applied for the enrollment. Two patients could not be analyzed due to a language error (protocols saved

Editorial Comment
The perioperative burden of hypotension is associated with several adverse postoperative outcomes. This retrospective study compared clinical responses and outcomes related to the prediction of upcoming hypotension using arterial waveform analysis data versus assessment based on the Hypotension Prediction Index (HPI). Cases with the HPI platform had a lower burden of hypotension. This suggests that this assessment tool might support improvement in the identification and early management of perioperative hypotension.
in German language could not be analyzed with the software at that time), so that two more patients were included to achieve 50 patients in both groups. All monitored patients matched the in-and exclusion criteria, no further patients were excluded.
Anesthetists were not aware of being part of a study to reflect real-life routine practice conditions. In-house treatment standard protocols postulate a mandatory MAP of at least 65 mmHg in all patients. However, to follow the requirements of the internal standard operation protocols, the individualized lowest blood pressure thresholds were kept higher in patients with corresponding comorbidities in order to take autoregulatory mechanisms into account. For all patients receiving extended hemodynamic monitoring a GDT protocol was mandatory to ensure a standardized but individualized treatment. As part of the GDT protocol, patients received crystalloid fluids and vasopressor therapy; inotropic medication and colloids were applied upon request of the clinician. In accordance with an ongoing multicenter trial 18 we used for FloTrac monitoring a mandatory MAP ≥65 mmHg but also a standard GDT protocol to optimize the cardiac output and oxygen supply. 19,20 In the HPI group, we used a personalized treatment protocol adapted from Maheshwari et al., 21 which simplifies the complex interpretation and the distinction in preload, cardiac contractility, and afterload as possible causes for hypotension to reflect flow and pressure. The GDT protocols suggest a therapy based on the current hemodynamics; hence the standard therapy was personalized for every (imminent) hypotensive event.
The HPI itself is a number on a scale from 0 to 100, symbolizing the risk for an imminent hypotensive event. Zero indicates the lowest and 100 the highest presentable risk. HPI alarm limits are not adjustable by the user, consequently a combined acoustical und visual alarm appeared when the index exceeded 85. 22 At that point, the clinician was instructed to review the patient's hemodynamics using the secondary screen ( Figure 1

| Data collection
In all patients, the arterial line was placed after anesthesia induction in the radial artery preferably at the nondominant extremity as precaution in case of catheter-associated complications. The advanced hemodynamic monitoring was started immediately after insertion and the quality of the waveform signals was continuously visually monitored by an experienced senior anesthetist.
We used the Edwards monitoring for the data collection: The baseline arterial waveform analysis was performed identically in both groups. Sample frequency was 100 hertz and the arterial blood pressure as well as the specific hemodynamic calculations were displayed continuously on and recorded every 20 s with the current HemoSphere advanced monitoring platform (Software release SHM 2.0.0.117, Edwards Lifesciences Corp.). 15 The HPI algorithm indicates the so-called extra parameters prediction index, systolic slope, and dynamic arterial elastance. 22

| Statistical analysis
The data analysis and statistical plan were written and posted on a publicly accessible server before data were accessed. Data col-

| RE SULTS
One hundred consecutive patients were included in the analysis from 5 September 2019 to 5 August 2020 (FloTrac sensor n = 50, HPI sensor n = 50). Epidemiological data were comparable in both groups and showed no significant differences in their distribution ( Table 1) Figure 4B).

| DISCUSS ION
Intraoperative hypotension in patients undergoing noncardiac surgery remains an important risk for adverse events. Ahuja et al. 8 recently provided evidence that the risk for myocardial and kidney injury increases with duration and severity of intraoperative hypotension.
In this study, we compared two systems of hemodynamic monitoring, which differ in a few calculation details while the analysis algorithm of the arterial waveform is identical. The HPI sensor has been assumed to be superior by calculating the prediction index of hypotension and granting additional information to enable the distinction in preload, cardiac contractility and afterload as potential causes for hypotention. 21   Our data suggest that aside from a study setting, using the advanced hemodynamic monitoring of the HPI seems to be a useful additional tool to increase our vigilance to hemodynamics since our focus is often shifted to the manifold requirements of patient care.

| Limitations
Certain restrictions are inherent to retrospective studies. Due to the design, a blinding or randomization was not possible, and con- The minimally invasive hemodynamic monitoring is only a mathematical analysis, the calculations are highly depending on the quality of the arterial waveform signal, making it unreliable with weaker signals. 36,37 Published data conclude that the different invasive and noninvasive hemodynamic measurements only have poor accordance. [38][39][40] Nonetheless, the HPI is able to predict hypotension in the presence of vasoactive and inotropic substances, but surgical causes of hypotension (e.g., clamping of the inferior vena cava) cannot be predicted because of the lack of physiological prodromal changes in the arterial waveform.
In summary, in this single-center retrospective observational study, we describe a significant reduction in incidence and duration as well as the severity of intraoperative hypotension for patients treated with the HPI compared to arterial waveform analysis alone. Our study thereby provides further evidence of the advanced HPI algorithm in clinical application enabling the transition from prediction to actual prevention of hypotension. To identify which patients benefit most from the advanced monitoring, we suggest to further investigate the demonstrated effects and their influence on the outcome of different patient populations conducting large randomized controlled trials.

FU N D I N G S TATEM ENT
The authors acknowledge that support for this study was provided solely by institutional and/or departmental sources'.